DESCRIPTION: (Adapted from the applicant s description) The investigators propose two classes of random effects regression models for handling multi-level clustering which occurs in studies of outcome in infertility, particularly those related to treatments by in vitro fertilization and embryo transfer (IVF-ET). Data from ongoing outcomes research on IVF-ET will be analyzed and used to study properties of the models. Two are studies of clinical factors which affect outcome, and the third uses digital image processing and video microscopy to grade embryo viability and relate it to eventual pregnancy status. The growing use of IVF-ET and other methods of assisted reproduction has raised a number of public health, economic, and ethical concerns. There is a recognized need to balance the success of ART with judicious application; this requires knowledge of factors which can optimize healthy pregnancies of low order.